ASSESSING ROOF-BOLT PERFORMANCE IN DEEP COVER: Using field results, NIOSH researchers calibrate models to determine the reaction of different bolt lengths.
To better understand load shedding and stress transfer on coal pillars due to retreat mining in deep cover panels, researchers from the National Institute for Occupational Safety and Health (NIOSH) conducted a monitoring field study. Two sites, at overburden depths of 1,000 ft (304.8 m) and 1,500 ft (457.3 m), were selected in a room-and-pillar mine in southern West Virginia, operating in the Lower War Eagle (LWE) seam. They monitored the deformation and stress changes in the roof and two adjacent pillars at each site during the retreat mining process. The results along with field observations were used to calibrate large-scale models for each site, which were used to assess different sizes of roof bolts.
During pillar retreat mining, the likelihood of instabilities increases due to elevated stress levels near the pillar line because of the abutment loading. During the last 10 years, about one-third of the ground fall fatalities in the United States occurred during retreat mining.
One interesting piece of information about pillar-recovery fatalities was that the victim was nearly always under bolted roof. Sheared and broken No. 5 (5/8-in.), fully grouted rebar-bolt failure contributed to three of four fatal roof-fall incidents that have occurred in deep-cover retreat mines.
This claim raised the following question at the study mine: Is it safe to use a 4-ft-long No. 5 rebar bolt (grade 60) at a depth of cover more than 1,000 ft? To answer that question and others related to retreat mining in deep cover panels, NIOSH researchers monitored the roof and rib deformation and the change in vertical pressure within two adjacent pillars during pillar recovery. Site-1 represents the monitored site at 1,000 ft of cover, while site-2 represents the monitored site at 1,500 ft of cover.
For this study, large-scale FLAC3D models were calibrated based on both field observations and instrumentation results. FLAC3D is numerical modeling software used for the geotechnical analyses of soil, rock, groundwater, constructs and ground support. The calibrated models were then used to compare the ground response and the induced stresses of two pillars at instrumented sites 1 and 2. The maximum lateral displacement (roof shift) obtained from the calibrated large-scale FLAC3D models was applied to a small-scale FLAC3D bolted models. Then a safety factor for axial stresses of the simulated roof bolts was calculated at the deep cover sites.
The typical geology consists mainly of interbedded shale, sandy shale, silty shale and sandstone. Underground geotechnical data on the immediate roof rock and the coal bed were collected. The Unconfined Compressive Strength (UCS) of coal was an input parameter in the coal-mass model used to simulate the coal seam. The UCS of the shale, sandy shale, and sandstone were determined and the results were provided by the mine. The average UCS for the shale and sandy shale was 5,677 psi (39 MPa) and 9,455 psi (65 MPa), respectively, while the UCS for sandstone was 22,145 psi (152.7 MPa).
The mine produces bituminous coal from the LWE seam by the room-and-pillar retreat-mining method. Site-1 and site-2 were located in the No. 6 entry in the third left panel. The panel width was subcritical and consists of eight entries and barrier pillars between the subsequent panels. The dimensions of the pillars are approximately 53 ft (16.1 m) x 99 ft (30.1 m) rib-to-rib (R-R). The entries and crosscuts were about 20 ft (6 m) wide at the instrumented sites. Site-1 was located between crosscuts 9 and 10, while site-2 was located between crosscuts 21 and 22. The mining height is approximately 5.2 ft (1.6 m) at the instrumented sites. There was neither sudden topographic changes nor multiple seam interaction in the third left panel.
During retreat mining, two continuous miners (CMs) are used simultaneously to extract each row of pillars. In this method of pillar extraction, mining progresses from the center of the panel outward in a staggered fashion. This method results in leave blocks that are left unmined in the center of the panel rather than on the sides. The number of these blocks in the panel corresponds to a panel width providing adequate pillar stability for a given depth of cover. In this case, one line of pillars is left unmined in the center of the panel at site-1, while two lines of pillars are left unmined at the center of the panel at site-2. Figure 1a and b show the pillar layout and extraction sequence at site-1 and site-2, respectively. Barrier pillars are used to isolate the active panels from the previous gob.
Reducing the panel width by leaving unmined pillars at the center of the panel also reduces the elevated stresses of the abutment loading during pillar retreat mining, particularly under deeper cover. No slab cuts were taken from the barrier and a final stump was left unmined to provide roof support during pillar recovery. The size of the final stump was a minimum 6 ft (1.8 m) x 6 ft (1.8 m).
Six wood posts are spaced 3 ft to 4 ft apart across the entry to provide support before a pillar lift was started. An 8-ft-long (2.4 m) cable bolt was used as a supplemental support in the intersection areas, the spacing between the cable bolts is approximately 6 ft (1.8 m). The mine is using one row of 3-ft-long (0.9 m) No. 5 (5/8-in) fiberglass bolts to support the coal pillar ribs in areas greater than 1,000 ft of overburden. The spacing of rib bolts at site-2 is roughly 8 ft (2.4 m).
Underground visual observation was one of the primary techniques used to evaluate the stress and deformation levels at the instrumented sites. Observational data can provide information about large areas of a mine and, when combined with instrumentation results, can be used to validate large-scale FLAC3D models. When the pillar line was located at the instrumented sites, the immediate roof, floor, ribs, and the extent of roof caving were observed in the retreated panel.
At both site-1 and site-2, there were no signs of the following: excessive roof sagging, open fractures in the roof, roof cutters near the rib line, or outby roof failure. Also, there were no signs of excessive roof-bolt deformations, such as excessive bolt elongation or severely deformed bolt-plates. The floor was not visibly heaved, and it was in good condition. In other words, the structural integrity of the immediate roof and floor was sufficient to withstand development stresses and retreat mining-induced abutment stresses at both site-1 and site-2. The entry between the two instrumented pillars at each site was accessible and the ribs were in good condition with little rib spalling.
In addition to the visual observation and evaluation, several test holes were drilled in the roof near the instrumented sites to evaluate the immediate roof for bedding separations and the lateral displacement/roof shifting. No roof shifting was recorded at either site test holes. However, roof bedding separation was reported at site-1.
At site-1, the immediate roof caved right behind the pillar line with no signs of roof overhanging, and the gob generally formed quite rapidly. However, at site-2 prior to the removal of the data logging system, the immediate roof did not cave, and was overhanging for at least one break, approximately --100 ft (30.4 m) in by the instrumented pillars. The narrowness of the panel width for site-2 and the presence of sandstone and silty shale in the roof may have restricted caving.
Instruments have been used increasingly in mines to measure deformation, stress, strain and load.
The instruments included were one four-point roof extensometer (RXC), two six-point roof extensometers (RXL and RXR), one borehole multi-point rib extensometer (MPRX), six borehole pressure cells (BPCs), and three hollow inclusion cells (Hi-cells). The Hi-cells were installed in the roof near the rib at 45[degrees] angles (see Figure 2).
The pressure change in the BPCs represents the induced abutment load due to only pillar retreat mining. The maximum pressure change in the BPCs in pillar 1 was 2,752 psi (18.9 MPa). The strength and stiffness of the immediate and main roof beams may have inhibited gob formation and caving at site-2, which can have a significant impact on stress levels at the pillar line.
The RXL and RXR measured the vertical displacement of the immediate roof. During retreat mining at site-1, the maximum and minimum anchor displacements were roughly 1.4 in (0.0355 m) and 0.6 in (0.0152 m). The RXR extensometer data was corrupted, as it neither matches the results of the RXL nor the visual observations at the site. At site-2, the maximum and minimum anchor displacements were roughly 1.2 in (0.0304 m) and 0.6 in (0.0152 m), respectively.
The two main factors that lead to an increase risk of rib falls during retreat mining are thicker coal seams and higher stress levels. The MPRX were used to monitor the deformation of the rib for pillar 2. The MPRX, however, malfunctioned.
Large-scale Model Setup
The visual observations and field instrumentation data collected at each site was used to calibrate the large-scale FLAC3D models.
The large-scale FLAC3D model was calibrated based on the following:
1) Comparing the vertical stress from the BPCs inside instrumented pillars 1 and 2 at site-1 and site-2 with FLAC3D model results. The calculated vertical stress from the model shows good agreement with the monitored vertical pressure from the BPCs for pillar 1 and pillar 2 at both sites.
2) Comparing the rib sloughing in the FLAC3D model with visual observations. The observed rib sloughing at the instrumented site-1 was less than 1 ft, while for site-2 it did not exceed 2 ft at the side of the pillar. Generally speaking, the FLAC3D models provide acceptable agreement with the observed rib sloughing at site-1 and site-2.
3) Comparing the vertical roof displacement in FLAC3D with visual observation and instrumentation results. The visual observation showed small roof deformations at both site-1 and site-2. Based on the instrumentation results, the average vertical movement of the immediate roof due to pillar retreat was approximately 1 in. and 0.9 in. at site-1 and site-2, respectively. The FLAC3D model shows that the roofline moved approximately 0.22 in. and 0.72 in. at site-1 and site-2, respectively, after deducting the vertical displacement due to development loading.
Maximum Roof Shift
Currently, the mine is using five 4-ft (1.2-m) No. 6 (6/8-in.), grade 60, fully grouted, nontensioned rebar bolt per row as the primary roof support in areas with a depth of cover greater than 1,000 ft, while in shallower areas the same configuration with No. 5 (5/8-in) rebar is used. The mine is considering using the No. 5 rebar bolt at both deep and shallow cover areas.
The first question usually asked about roof bolts is: "What is its capacity?" Two types of capacities are known for roof bolts: yield and ultimate. These can be calculated from the diameter of the bolt and the properties of the steel. The yield and ultimate strengths of grade 60 steel is 60,000 psi (413 MPa) and 90,000 psi (620 MPa), respectively. In this study, since there is some uncertainty about the ground conditions/roof properties in the bolted horizon in addition to the uncertainty about the loading conditions due to variable gob caving characteristics, it is prudent to design the capacity of roof bolt systems based on the yield strength rather than the ultimate strength.
A small-scale FLAC3D model was generated to calculate the induced axial stresses in the 4-ft No. 5 and 4-ft No. 6 fully grouted rebar bolts when they are subjected to both an axial load of 2.9 tons and a roof shift of 1/8 in. The 2.9-ton vertical load was used to simulate a detached rock-block from the immediate roof of size 3.5- x 3.5 x 3-ft, assuming the unit weight of the rock is 159 lb/[ft.sup.3]. The small-scale FLAC3D model consists of two identical rock-blocks of size 4-ft x 4-ft x 4-ft.
The model was solved in two steps: in the first step, the bolt was pulled by a 2.9-ton vertical load applied on the bottom block. In the second step, a horizontal displacement of 1/8 in (3.2 mm) was applied on the lower block to simulate the roof shifting. The induced axial stresses after the second stage of loading in the No. 5 (5/8-in) and No. 6 (6/8-in) bolts due to vertical and horizontal movements are obtained from FLAC3D models.
Based on the model results, the safety factor based on the yield strength for the No. 6 (6/8-in) rebar bolt is 2.47. Based on the two calibrated small-scale FLAC3D models, the No. 5 rebar should not be used if the depth of cover exceeds 1,000 ft.
This article was adapted from a paper delivered at the 2019 Annual Society for Mining, Metallurgy and Exploration meeting, A Case-Study of Roof Support Alternatives for Deep Cover Room-and-Pillar Retreat Mining Using In-situ Monitoring and Numerical Modeling, by G. Rashed, CDC NIOSH, Pittsburgh, Pennsylvania; M. Sears, CDC NIOSH, Pittsburgh; J. Addis, CDC NIOSH, Pittsburgh; K. Mohamed, CDC NIOSH, Pittsburgh; and J. Wickline, Coronado Coal, Charleston, West Virginia.
Caption: Figure 1--Pillar retreat plan for a) site-1 and b) site-2. Instrumented sites are marked by red circles.
Caption: Figure 2--Instrument type and location for both site-1 and site-2.
Caption: Figure 3--Loading conditions and pillar line locations at the two instrumented sites.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||ROOF BOLTING|
|Publication:||Coal Age (1996)|
|Date:||Mar 1, 2019|
|Previous Article:||KENTUCKY PILOT PLANT PRODUCES RARE EARTHS, CONTROVERSY: Extracting rare earth elements from acid waste is proven to be technically feasible in...|
|Next Article:||IMPROVED LOADING TECHNIQUES THROUGH SITUATIONAL AWARENESS.|